For the last 20 years our research group has studied extracorporeal circulation, with the intent of developing prolonged extracorporeal life support for acute respiratory failure. For the past 16 years this work has been supported by NIH. As a result of these studies extracorporeal membrane oxygenation (ECMO) has been studied in the laboratory, evaluated in adults, and has become standard treatment for respiratory failure in newborn infants which is unresponsive to other forms of therapy. There are now 60 neonatal ECMO centers worldwide, each attaining survival rates of 80-95% in moribund newborn infants. We are currently concluding a Phase III prospective randomized study to evaluate cost effectiveness and morbidity of ECMO begun early in the clinical course compared to conventional ECMO and conventional therapy. Having established that ECMO is not only feasible but standard therapy in one group of patients, we now propose to improve and simplify the technique, then extend clinical trials to premature infants and older children with respiratory failure, based on the following rationale. The major problems limiting further application are bleeding, vascular access, complexity, and irreversible lung damage. The first three problems are solvable by laboratory experimentation, and we have completed pilot studies indicating that the elimination of heparin, single catheter venovenous access, and total servoregulation are now possible. Armed with this improved technology, we propose to return to clinical trials in premature infants (where intracranial bleeding is the current limitation), and older children (where reversible lung damage is the current limitation). We propose to follow this rationale for the next five years by pursuing four research projects simultaneously: Technology, trials in term infants, trials in premature infants, and trials in older children.